Project Description

Smart(er) cooking

PixFood is innovating the way we interact with food by using AI and image recognition to deliver the right meal at the right time.  

About the client

IMP, a well-established media publisher, was facing the challenges of the digital age. They had to find a way to adapt their recipe publishing business to the ever-increasing needs of Millennials, the target market. 

So, we had to find a way to provide customers with hyper-personalised recipes that are quick & easy to cook, healthy, and most importantly, delicious! 

The AI-driven food assistant 

Through in-depth research & market analysis, we came up with the concept for a digital food assistant that uses advanced image recognition and machine learning to provide the best recipe choices for millennials.  

And so, PixFood was born – a digital companion that understands the context of users’ everyday life and habits 

The simple & fun user experience starts with the user taking a photo of any ingredient they want to cook and getting tailored recipe suggestions that also take into context the time of day, the season, and your dietary preferences.  

For example, if you open the app at 6 PM on a Monday in winter, it will recommend a dinner type meal that has the ingredient you took a photo of, is a common wintery meal, AND fits your vegetarian diet! 

All of the recipes in the app are carefully selected and they’re featured with high-quality step-by-step videos created by professional chefs and produced in-house.  

PixFood - Smart Food Assistant

The technology behind it

The PixFood digital platform was built entirely on AWS Cloud infrastructure, enabling us almost limitless scalability and the option to add new features on the go.

The app was packed with high-quality video recipes, so in order to deliver these in an optimised & fast way – we used Lambda as a serverless computer, S3 hosting as an admin panel for static sites, and then CMS and Storage Asset to compress the videos when uploading, thus creating different formats for each video.

All data from the image recognition feature was stored in Dynamo DB and then used for in-depth analysis and to power up the algorithm for user recommendations.

Finally, we implemented Mixpanel’s product analytics platform on top to create customer segments based on real-time behaviour, and deliver highly personalised messages to them via email, in-app messages & push notifications.

Benefits of using AWS cloud infrastructure

  • Scalable – can serve 10,000 or 100 million+ users

  • Fast – hosted close to users for fastest response time

  • Expandable easily add or remove features

  • Flexible costs are automatically adjusted to usage

  • Always on zero downtime

PixFood - Mobile App Features

The outcome 

Our goal was to improve millennials’ everyday life by providing them with tailor-made recipes that fit their food preferences. 

 And we did it with PixFood – the app that gave a totally new approach like no other food app on the market. The word quickly spread, grabbing international recognition.  

  • PR mentions in TechCrunch, Entrepreneur, Lifehacker and more

  • Nr.3 product of the day at Product Hunt

  • 0.5$ minimal cost-per-install from paid acquisition, racking up 30,000 downloads in the first 3 months

  • Fully scalable solution that can support 10k to 10M users in minutes

  • 1M+ personalized emails, push & web messages sent

  • 2,500+ recipes cooked monthly

Got a project
we can help with?

Get in touch with sales